[R-sig-Geo] variogram model with gstat

[Edzer Pebesma]

Try:

library(sp)
demo(meuse,ask=F)
library(gstat)
m = meuse$zinc
meuse$m = (meuse$zinc - mean(meuse$zinc))/sd(meuse$zinc)
var(meuse$m)
# [1] 1
variogram(m~1, meuse, cutoff=1e6)$gamma
# [1] 1

you see that the semivariogram averaged over ALL distances equals the
variance of the variable. What you look at is (for good reasons) the
semivariogram computed up to a certain distance.

Hth,

On 12/11/2013 06:17 PM, Moshood Agba Bakare wrote:
> Dear All,
> I used gstat package of R to fit variogram model to experimental variogram
> having normalized the raw data as
> 
> VN = (VP -VMean)/VSd
> 
> Where VN is the normalized yield value, VP is the original yield value at
> sampling point, VMean is average yield  value, and VSd is the standard
> deviation of yield value.
> 
> The reason for normalization is to be able to assess the stability of
> spatial pattern of yield over a number of years. Having normalized the
> data, I expected the sill i.e. the variance of the data to be one but to my
> surprise the semivariogram plot shows sill of 0.8. What is responsible for
> this?
> 
> The result of fitting exponential function showed a nugget effect of zero
> as showed below
> 
> ## Exponential variogram model> exp.var<-vgm(psill = 0.8, model="Exp",
> range= 15, nugget= 0.1)> exp.mod<-fit.variogram(empvar,model=exp.var)>
> print(exp.mod)  model     psill    range
> 1   Nug 0.0000000  0.00000
> 2   Exp 0.8066099 11.36279
> 
> 
> Does it mean the exponential structure explained 100 percent of total
> variation in a normalized yield variable? I am confused! Have a look at the
> attached file for the semivariogram plot. Do I need to detrend before
> fitting the model? it seems the plot increases monotonically without level
> off. Please advise.
> 
> Thank you
> Moshood
> 

-- 
Edzer Pebesma
Institute for Geoinformatics (ifgi), University of Münster
Heisenbergstraße 2, 48149 Münster, Germany. Phone: +49 251
83 33081 http://ifgi.uni-muenster.de GPG key ID 0xAC227795

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 555 bytes
Desc: OpenPGP digital signature
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20131211/9cf448d0/attachment.bin>